AI-GENERATED DIETARY GUIDANCE FOR CHRONIC KIDNEY DISEASE: A COMPARATIVE EVALUATION OF CHATGPT, COPILOT, AND GEMINI

Abstract

Background: Chronic kidney disease (CKD) requires strict dietary management tailored to disease stage and individual needs. Recent advances in artificial intelligence (AI) have introduced chatbot-based tools capable of generating dietary recommendations. However, their accuracy, personalization, and practical applicability in clinical nutrition remain largely unvalidated, particularly in non-Western settings. Objective: This study aimed to comparatively evaluate the dietary recommendations generated by three leading AI chatbots: GPT-4 (OpenAI), Gemini (Google), and Copilot (Microsoft), for patients with CKD stages 1 to 5. Methods: Standardized patient profiles representing each CKD stage were developed and used to prompt each AI model with the same culturally adapted request for meal planning. AI-generated diets were evaluated by three physicians using a 5-point Likert scale across three criteria: personalization, consistency with guidelines, practicality and availability. Descriptive statistics, Kruskal–Wallis tests, and Dunn’s post hoc tests were performed to compare model performance. Additionally, a detailed nutritional analysis of four meal plans (Initial, GPT-4, Gemini, Copilot) was conducted for Stage 3 CKD using both GPT-4 estimates and manual calculations validated against clinical dietary sources. Results: Gemini achieved the highest overall scores, particularly in personalization (mean = 3.91), followed by GPT-4 and Copilot. Statistically significant differences were observed across AI models in personalization (p = 0.0001) and consistency (p = 0.0002), with marginal significance in practicality (p = 0.0476). Nutritional component analysis revealed discrepancies between GPT-4’s internal estimations and manual values, with occasional deviations from clinical guidelines, especially for protein and phosphorus intake. Conclusion: While AI chatbots show promise in delivering dietary guidance for CKD patients, substantial variability exists across models in terms of personalization and clinical accuracy. Gemini demonstrated the strongest performance overall, but caution is warranted in clinical use, especially where nutrient estimation reliability and guideline adherence are critical. Further development and clinical validation are needed before AI-driven tools can be fully integrated into patient-centered CKD nutritional care.